CVE-2026-99595: sched/fair: Clear throttle state when leaving the fair class
In the Linux kernel, the following vulnerability has been resolved:
sched/fair: Clear throttle state when leaving the fair class
Task-based CFS bandwidth throttling removes a task's fair schedentity while the scheduler core still considers the task queued. p->throttled carries the fair-specific state across dequeue/enqueue operations.
DEQUEUECLASS removes the task from cfsrq->throttledlimbolist but retains the bit. If the last finite CFS quota is disabled while the task runs in another class, it is absent from the unthrottle walk and the bit survives. Returning to fair while cfsbandwidthused() is false then performs a real fair enqueue with p->throttled still set.
After bandwidth is re-enabled, a class-change dequeue treats the stale bit as limbo state and skips dequeuetask(). p->se.onrq therefore remains set, and putprevtaskfair() places the entity in the EEVDF tree while the task moves to RT. On exit, the fair picker can select the TASKDEAD task. schedule() then returns to dotaskdead(), which hits BUG().
Keep p->throttled across migrations and group moves, but clear it on DEQUEUECLASS because fair-specific state must not outlive fair-class membership. A later return to fair uses the normal enqueue path. That path still respects a throttled cfsrq.
A KASAN, LOCKDEP and DEBUGLIST v7.3 kernel reproduced the dotaskdead() BUG in 3/3 root and 3/3 delegated uid-1000 runs with CapPrm=CapEff=0 and RLIMITRTPRIO=1. The fix passed both matrices plus a 3/3 class-switch control with bandwidth continuously enabled, without a scheduler warning, KASAN report, lockdep splat or crash.
An AI coding assistant analyzed the state mismatch and drafted the initial reproducer, fix candidate and this changelog. The source paths, trigger, controls and results were rechecked against the cited revisions.
Affected Software
Remediation
Recommended actions to resolve this vulnerability, in priority order.
- Upgrade
Upgrade
linux/Kernelto a version that resolves this vulnerability.Fixed in 7.3.5 - Upgrade
Upgrade
debian/linuxto a version that resolves this vulnerability.Fixed in 6.1.176-1Fixed in 6.1.192-2Fixed in 6.12.107-1Fixed in 6.12.117-1Fixed in 7.3.5-1
Event History
Frequently Asked Questions
What conditions are needed to trigger the scheduler failure?
The issue requires task-based CFS bandwidth throttling, a task leaving the fair scheduling class while throttled, and the last finite CFS quota being disabled while that task runs in another class. The task must later return to fair scheduling, have bandwidth re-enabled, and undergo a class-change dequeue, allowing stale throttle state to leave its fair scheduling entity queued.
What is the observable impact if the stale state is reached?
A fair scheduler picker can select a TASK_DEAD task after its entity is incorrectly left in the EEVDF tree. The scheduler then reaches do_task_dead() and triggers BUG().
Does moving a task between CPUs or task groups clear the affected state?
No. The described fix intentionally preserves p->throttled across migrations and group moves. It clears the state specifically when DEQUEUE_CLASS removes the task from the fair class.